Optical Power Measurement
This domain covers absolute and relative optical power measurement: the power meter, its detector types and calibration, dBm versus dB, and the launch conditions that make readings meaningful. Almost every other test in this exam is built on a correct power reading.
dBm and dB
Power in dBm is 10·log10(P/1 mW). Memorize the ladder: 0 dBm = 1 mW, −10 dBm = 100 µW, −20 dBm = 10 µW, −30 dBm = 1 µW. The difference between two dBm readings is a loss or gain in dB. A meter that reads −10.0 dBm at the source and −22.4 dBm at the far end tells you the path lost 12.4 dB.
Choosing and setting the meter
- Detector material: silicon covers roughly 400–1100 nm (850 nm multimode); InGaAs covers 800–1700 nm and is the general telecom choice; germanium covers 800–1600 nm.
- Wavelength calibration: set the meter to the actual source wavelength. Each wavelength has its own calibration point; the wrong setting injects error, sometimes a lot.
- Adapters: 2.5 mm for SC/FC/ST, 1.25 mm for LC/MU, and keyed UPC versus APC. An APC ferrule forced into a UPC adapter gives a poor, unstable reading.
- Range: most meters saturate near +3 dBm; use an attenuator for higher powers.
Making a clean measurement
Inspect and clean both connectors. Let the source warm up and stabilize (15–30 minutes for a lab source). On multimode, control the launch with a mandrel wrap or an encircled-flux conditioner, otherwise the reading depends on how the light was launched rather than on the fiber. Take the reading, record the wavelength and meter serial number, and repeat if the value moves.
Worked example
A transmitter is specified at −5.0 to 0.0 dBm. The meter reads −7.2 dBm at 1310 nm. The technician cleans both connectors, re-inspects, and re-measures: −5.9 dBm. Reasoning: the 1.3 dB shift came from contamination, not from the transmitter; the reading now sits inside specification. Had it stayed at −7.2 dBm, the next checks would be the meter's wavelength setting, the reference jumper, and finally transmitter degradation.
Traps that cost marks
Reporting dBm when the answer is dB; using a meter set to 1310 nm to read a 1490 nm PON downstream; quoting '±5% of reading' as if it were ±0.2 dB; measuring an upstream PON burst with an ordinary meter (it averages the burst, so the reading is meaningless); forgetting that a dirty connector changes both loss and backreflection.
Key points
- dBm is absolute power; dB is the difference between two readings.
- Match the meter's wavelength setting to the source or the reading is wrong.
- Silicon suits 850 nm; InGaAs covers 800–1700 nm for telecom work.
- Multimode readings need a controlled launch (mandrel or encircled-flux source).
- Typical meter accuracy is ±0.2 dB, worst case around ±0.5 dB.